Marketing to Humans Changed

Thank you. Thank you. What AI really does to them? How does it change products, markets, customer behavior?

In my own work, I have looked at this topic since a long time. Since 28, I have been focused on how digital systems shape, who gets seen, who gets trusted, who gets chosen. I actually built a company called Fisher Analytics. This was early on in the social media area, social analytics platform.

I more recently looked, like, and built up a GEO measurement platform called QueryEdge, where we kind of, like, measure, like, how brands are displayed. As those large language models influence our world, as we do way more on OpenAI and Gemini, it is a question, who gets surfaced, who gets cited, who gets recommended? So it’s a new era in the face of the Internet, and the website in itself is no longer the one major point of entry. And therefore, it is super important for us to discuss how are agents shopping, how are agents perceiving our brands, and that’s exactly what we’re going to discuss today.

So I’m super excited that I’m joined by Alex Deas. He’s co-founder and CEO of Meridian, and we will discuss what marketing and brand leaders, need to understand on about GEO or the new SEO, how AI agents will shape the future of search and discovery. Alex, welcome to the show. Hey, Lutz.

Thanks for having me. Thank you. So what I would like to do, like, very often, it’s like people might not know you. Therefore, let’s do a quick intro.

Who is Alex? And tell us something about him. How you actually got to the place where you are today. Yeah, absolutely.

Well, like you mentioned, I’m one of the co-founders here at Meridian. We essentially build agents to help our partners market and sell to AI. So kind of agents that market to other agents or AI. I got super interested in this space by, I worked in like FinTech and embedded FinTech before this.

And I think kind of like the next frontier of FinTech kind of became this like agenda commerce route, right? How does the buying journey and the actual like checkout process change as decision making moves from humans to AI? Kind of the mental model that I came through is, you know, the AIs are essentially eating the entire buyer journey. I think simplistic mental model kind of be finding, evaluating, choosing and then buying products.

I mean, like the state of the world that we were at, AI was very much influencing those first three. Maybe not so much the agenda commerce part, which we can definitely dive into as well, but the actual buying. But in terms of, you know, like actually building solutions, pretty much every business doctor needed help understanding and improving around those first three. So I started Meridian about eight months ago and have been building ever since.

Very cool. How big are you guys? We’re a little under 10 employees right now. 10 employees.

Actively hiring in New York. So if you are an engineer, growth, really any specialist, we’d love to chat. Very cool. That is definitely the audience to talk to because we have a lot of visitors from New York.

Actually, let’s talk about this early on. And most of the folks who come regular to this channel know this, but we will take audience questions. So see the link below and like usual, put in your comments, put in your questions. If you place your name in it, I will call you out.

So off we go. Ask questions if you want to. Now, the main topic, here is websites used to be built for humans. It was a huge setup.

So websites were built for humans, but your next customer actually in today’s world may arrive through an AI agent. And those never miss any context. Why did you build Meridian in this world? Yeah, it’s a great question.

I think, you know, for me, it was really about the market and then this market is not even like, I think one is one of the most fascinating markets that we’ve ever seen. I think they’re really like the high level is there’s this shift happening in like decision making, especially for our customers in like purchasing decisions. Right. And this is shift quite simply from humans to artificial intelligence.

And there’s a lot of like different places that we can kind of like break that down to. But in a shift of decision making, kind of like this of this magnitude, which affects pretty much every industry, pretty much every business, any business that wants to be more scalable online. I think kind of diving into that was like an opportunity that I could not be a part of. So describe the shift a little bit.

So we have heard that SEO is about being found and that GEO is now being more or less about being the trusted and understood source for a machine. But explain that. So what is the shift? What does it mean?

Yeah, I think kind of the way I like to think about this is actually kind of like backing up like very quick, like history, right? Is, you know, when the Internet kind of started popping up 25 years ago, you had all these like searchers, right? And then on one side and you had kind of all these like end content or sellers or merchants or publishers on the other side. And it wasn’t a great way to connect those two.

And Google ended up building like a pretty awesome business connecting those two. But Google’s real, you know, model or KPIs, they wanted to get you from the search to the end content as fast as possible and actually away from Google. And that’s a big reason why they won the war, because they didn’t try it like Yahoo to keep you on Google for as long as possible to show you a bunch of ads. But in that three party model is kind of like disintegrating in real time.

You’re kind of getting into a two party model, which, yes, you know, consumers, businesses or personal might search in AI, but they’re also doing more and more of the entire buyer journey, right? They’re asking follow up questions, they’re asking what they should actually buy from. Like their use case. And when that happens, the third model or the actual like end business or their website, which was meant to convert and like follow that buyer journey is essentially getting pushed out, right?

So when we look at what these websites, what their main KPI is, it’s not take a search and convert them from the human web to actually like the demo link or to the buy or anything, because those decisions are getting made before the consumer actually hits the website. And so when we think about like, what is the purpose of websites today, it needs to be feed data into the website. And so we’re going to have to look at the AI where these customers are actually making those decisions. I’m talking quite often to companies who then tell me that their search traffic is declining, meaning the homepage is no longer the front door anymore.

The model is essentially. And what you are describing is that it’s not just a decline of the traffic to that front page. It is actually the whole. The.

decision-making process is happening before. At which point does it actually go into the website? Yeah. I think in terms of like when it goes into the website, it kind of does depend a little bit on like the use case for those companies.

We are seeing this rise, like you mentioned of zero click, right? Which is the idea that consumers are getting the questions or the answers to their questions without having to click a link. And we do expect really like for these large publishers, for these large companies, search traffic will generally be going down probably for the rest of the time. Now, the people that actually end up on their website are super high convert, right?

You see like a chat GPT referral is about 20 X more likely to a Google search because they’ve already made the decision. They’ve already gone through that buyer journey. But it very much is, you know, when we think about like, what are the KPIs that we work with on our customers? It’s no longer just overall total amount of search volume.

You’re kind of like shooting yourself in a foot and going against the market. What you really want to be figuring out is how do I get the most high intent as many of those high intent consumers that are actually going to convert to my website versus just overall, you know, how many organic searches am I getting? But that’s an interesting discussion because now we are still keeping the same user journey. If I see you correct, right?

Before we had, we tried to get many people into the top of our funnel on our website and we wanted to convert them. And now you’re saying we still want many people on our website, but the intent is changing. Do if I’m a brand, can I actually assume that actually before we go down there, like, let’s ask a different question. If I’m a brand, what’s your highest, like high value pitch to them?

What should they do? And then we break it down into the user flow. Actually, let’s start there. If I’m a brand, what do I get right or wrong?

Um, it’s a, it’s a good question. That’s a complicated question. Right. Um, but you know, the way we kind of think about it is like you mentioned like, yes, we want a lot of people on our website as much as possible, but if that is the core battle that we were fighting, we were fighting a losing battle, right?

Zero click is only going to, uh, continue to kind of move into like every search in our base that we possibly have. Um, and so when we think about like, what should we actually be focused on and the KPIs, it is very much like we want to emulate what our buyer is. We want to understand how LLMs are trusting, uh, product solution services for the markets that we are going after. And then we want to make sure that we are doing the things that, uh, actually help those decisions.

It isn’t very much, you know, maybe like the rough funnel is the same, but the part of that funnel that we own is much smaller and much more at the end of it. Right. Um, we own the other parts, but just not. Like the ways that we would like, it’s not on our website.

We might be feeding data to like end up in one of those AI searches. Um, but we need to be realistic. I think about like, uh, where this industry is going and what our website purpose is. And then how do we like actually draw KPIs and like make sure that we’re measuring like the right things for our business.

Totally. Super fascinating. The conversion has changed. And this is something which many executives don’t yet understand.

Um, we did some research, uh, with, uh, my team where we try to emulate a discussion like chat GPT on the website. If you do that, the conversion goes way higher, meaning now a company is see less people coming into the funnel, but they’re converting stronger. Is that what you see with your customers as well? You know, it really, I think like for some industries, yes.

For some industries, no. Right. Uh, anything that can be directly purchased online, I think more of like the e-commerce, the e-commerce type of companies, um, they are going through a very different type of, uh, type of change because everything that can be purchased online will soon be purchased in like, Hey, chatbots are directly by the agents. Right.

I think the industry is like maybe a B2B SaaS, you know, book a demo, reach out, let’s hop on a call, uh, a little bit going to be, you know, a couple of years behind that. So the short answer I think is like, you know, for some industries is happening like right now and it’s kind of like an edge to edge hold crisis. Um, for other industries, they’re holding a little bit more of that, like, bottom of funnel for now. Um, but it is kind of affecting really each part of every industry.

Got it. Got it. And, um, what’s a good framework for an executive to think about when they kind of are faced with, um, this change. So essentially what you’re saying, SEO is changing over to geo, geo is generative engine optimization.

How do you approach it now? Yeah. I think there’s like two. Yeah.

Kind of like very high level frameworks that I like to think about. One is more of just like an attitude around marketing, which is your primary marketing should be, you are now marketing towards computers versus humans. And that is very different, right? Computers are a little bit less emotional than humans.

They make decisions a little bit more, um, probably smartly kind of like humans, humans make it. But when we’re thinking about like, you know, how do we invest these dollars into marketing? Uh, we should be thinking about how do we make sure that we are marketing to the machines, uh, uh, while also, you know, marketing too. To humans.

Uh, so I think that’s like very much how I like to frame the problem, like the shift for these companies and the executives around like, what is the high level of it? Perfect. Yeah. Maybe let’s do, and this is like normally in this, um, um, live show, we, um, don’t really do show and tell, I would like to show how we could order in the future, um, something.

And one, There’s also these, obviously, to Moot Cuff Coffee, and I can read about where they have what type of coffee, and I can learn about the stories about their coffee. I can go and shop all coffees. I can look at my medium, whatever roast, and trying to understand which coffee is the right one for me. And that would be the normal customer journey.

Now, I as well, as you can see, I have Claude. Claude is, I use it as a plugin onto my Chrome, and all what I can now do is I can say something like, I like fruity and experimental coffee flavors. I would like to order $100 worth of coffee. If I do that, then now Claude will make, create a plan, browse the website, find fruity and experimental options, select the coffee that matches this preference, and put it, add it to the chart.

And if I approve, then what you will see is it will actually now create, it has created the plan. It creates now the screenshot. It goes to the products. It’s probably on the wrong page by now.

So it will read the page and will try to, again, to go to different options. You see that it’s operating the website. It goes now through every part of the website. It reads everything about every coffee and then decides based on the input, experimental coffee flavors, it starts to adding them in.

So at the moment, it’s just reading all the websites and then it will add it in to my bucket. If you are Mort Cup Coffee, you will have no clue why there is one user who reads all the pages, that this is not a real user. This is just an AI agent from Lutz Wenger. And you see now it’s actually putting Disco Peach Connection in.

Interesting coffee. Here we go. And it will do it now up to $100 and then it will stop and tell me I succeeded. Okay.

So we can stop this view here. Alex, in such a world, how does Mort Cup Coffee, how would they deal with it? Yeah, it’s a great question. Of course, there is no…

There’s no need for a website anymore, essentially, right? I think I would switch the framing there a couple of things, right? You kind of describe this as not a real user, but I do think this brand should be thinking about this as the realest user that they’ll have for the next 100 years. This is their user now.

These things, maybe I’m wrong, but I do think they’re generally going to only get better and more prevalent. And 99.9% of the things visiting their website will be these users, right? Yeah. And it’s very authentic.

Very well said. This is the user. So the AI agent is the new user and not the human anymore. Yeah, I think so.

And it’s a great experience for you, right? You don’t have to do all of that like clicking around anymore. You get totally abstracted. But I think the second thing there is like, it’s not that there isn’t a…

There absolutely is use for websites, right? These are large language models. They are trained and they know how to operate the internet. Yeah.

There’s just like five or five hundred I don’t know, I think a lot of the tactics that we used to use for humans, for websites, think about, are they actually the best way to market our products, services, solutions to machines? And if not, might need to update our websites a little bit. When you say AI is reading your website, what does it mean for a brand in terms of what content will work and how they need to update the content? Yeah, I think there’s really two things to take away from here.

One, websites were not designed as clean data delivery mechanisms to computers. They were designed to appeal to you or I, be able to look at that website. Hopefully, it’s pretty as well. And I love pretty websites.

We have a pretty website, probably does a little bit more for the human than the machine. But we need to be thinking about one, okay, can machines easily… understand and cite my content. And then two, the things that are on your website really should be as close to the source of truth about your product, solution, service, company as possible.

If you sell to, I don’t know, maybe 20 industries, and you used to think that that was implied by your website, you should be very explicit on your website about who you sell to, who your target use case is, who your pain points are, and follow that entire buyer journey and make sure that it is in language. Very explicit for that AI, because if it’s not, then the AI has no way to know, and they might recommend somebody else who is a little bit more explicit in what they do or who they solve. We used to say that being not explicit is the kryptonite for a PM, right? If there’s ambiguity as a kryptonite for a PM.

Can we say the same approach is true for marketing? Meaning you have to be specific? Yeah. I think that’s a good question.

The more specific you are, the better it is? Yes, I think absolutely. We want to try and mirror that buyer journey as much as possible, to give a quick example. It’s no longer best CRM typed into Google search, and then you read a bunch of blogs around the different types of CRMs for different types of companies.

It is very much the user or the buyer might be putting 100 characters or 100 words into the chatbot around, I run a small business in Idaho. I have these seven, things that I’m looking for in a CRM, go find me the right CRM. And if you are that CRM company and you do not have answers to those seven explicit things, it’s very hard for the AI to recommend whether you are or aren’t a good answer to that question. So we do want to be very explicit.

And you can be broad explicit, right? You can say we sell to everyone, but you should say who you do or who you don’t. Now comes an interesting question. We just looked at the coffee example, and they have coffee tastes like, by the way, I have no association with MootCount, just saying, just an example.

They have tastes like peach, for example. Now, how does an LLM understand that peach is experimental or that peach is outside of the norm? And how do you as a marketeer have to explain this? It’s a good question because does AI understand it?

I think to some degree, maybe, but I don’t think to the degree that you or I might do it. So I do think there is this kind of different way that we use words or language that maybe an AI can or can’t understand better than a human. To your explicit answer, it might understand for that wording. It might not.

But that is the truth about how these large language models work. They are, at the end of the day, like next token prediction machines. Obviously, we’ve seen a lot of progress with reasoning and making sure that those predictions are grounded more in truth, especially in the latest few model generations. But I think it’s a great question around how are we framing our products and services in language that is a little bit more easier for AI to understand, like human.

Which is, by the way, the interesting part for placement of a brand within a company. So I think that’s a great question. That’s the reason why I brought up peach. If you think about a typical coffee, it’s nutty, chocolatey.

This is the typical Italian-French coffee style. Meaning if I put on my website, my coffee is chocolatey, that is normal. Like most people would say that. And therefore, it’s not very special for an LLM to learn this.

And therefore, an LLM wouldn’t pick it up. If I say my coffee tastes like peach, LLMs do not have peach taste as a common taste for a coffee. So the more novel a concept is for an LLM, the more it will be picked up. Does an LLM understand what peach taste is?

No, not at all. But it does understand that most coffees taste like chocolate. And here something tastes peach. So this is interesting enough to keep the information.

So this is how LLMs learn. What is, how can I use that as a marketeer now? Yeah, it’s a good question. You know, I think, you know, I don’t want to say that this is like a, the death of like, brand marketing or inspiring marketing, right?

But it is a little bit more like we need to be a little bit more to the point with our marketing. You know, these like long, seven paragraph, maybe six paragraph, you know, like, like, stories about like the history of a coffee is probably a little bit less likely to appeal to an LLM who’s looking for like a very specific set of characteristics for like a user journey, then maybe the human that was actually like reading that page in the first place, or the, you know, SEO copywriter who was trying to stuff that with like keywords. And so, you know, to answer your question is like, what is the goal, I think, of this, like coffee brand, like maybe if they can use a little bit more of that, like peachy language, if their goal is to be, you know, recommended for, you know, what is the goal of this coffee brand? Like, maybe if they can use a little bit more of that, like peachy language, if their goal is to be, you know, recommended for, you know, what is different or user queries of like, you know, what is very different type of like coffee that I should try?

There’s probably a very large amount of queries that are like what are the best, you know, coffees or the best XYZ coffees, and they’re probably typing in very specific known coffee flavors, right? And so I think it is like what is the what is the goal of that marketer, but making sure that you know, the language they use is very reflective of the types of query queries and then type of machine that they’re looking for. There’s also like, of who their ideal customers are actually searching. So specificity is a big topic, and I hear you loud and clear, meaning as a marketeer, we should be specific to the questions our customers ask and not a lot of stuff around it, like coffee history, where did you get it, what is it?

Be specific, totally get it. If I want to be seen in this world, I would claim novelty as well is a big topic. Do you see that as well? It’s not a super straightforward answer.

I think it kind of depends, but I don’t think just specifically being novel will give more likely of an AI to rank you up. Of course, you need it to be differentiated, but not novel. I think differentiated really comes from authority and comes from research. There’s a lot of things that AI does care about.

It isn’t just the word. It’s can you back up the words that you were saying with essentially proof on the internet that what you’re doing is correct. In terms of consumer products, that’s very often authority, refuse, that kind of thing. But I think we shouldn’t be using, I wouldn’t say for these customers, I wouldn’t recommend using differentiated language that might be very different than what their actual personal language is.

I think that’s a very specific thing that a potential customer is searching just for the reason of using differentiated or if it’s differentiated, but true and very specific to what the product is, absolutely. But I don’t think that I’ll be like, we want to make sure really honestly, it’s question and answer. AI chat is question and answer. We should be thinking about all of our marketing in terms of answering those questions for our customers.

Wow. There’s a lot to unpack. There’s three areas you touched on. One is be differentiated.

One is be trustworthy. Like have authority in where you are differentiated. And the third one is how do you communicate, be specific in question and answers, right? These are three different topics.

Yeah. If we go to the peach coffee, then obviously if the coffee doesn’t taste like peach, it’s not a good idea to say it’s a peach coffee. Who knows whether people want to like peach flavored coffee. But Alex, I tested it.

It’s amazing. Anyway, who wants peach flavored coffee? Probably not most people. And therefore, having this differentiated claim might, if you can’t say it with authority, then it won’t work.

Totally get it. However, if you can differentiate yourself from the general knowledge base, then there is an easy way. There is an easier way to be seen, I would claim. Does it like, is this, do you see that, do you see those structures as well?

Yeah. I think that makes total sense. Right. I do think, you know, I am kind of, I think like, repeat this a lot around like this, you know, what I work with, like our executives and our team and our customers along around like, you know, mirroring those questions.

I think in terms of like the coffee example, right? Like, I don’t know. I don’t know kind of the breakdown of like how many people are searching for like peach coffee. Right.

But if you wanted to add language around like, hey, what is a great gift for my mom for Mother’s Day? Or what is a great, you know, like gift around like holidays. And then you’re looking at like, okay, I’m trying to like offer a differentiated product that is specific to kind of like an intent around when somebody would want to buy a differentiated product. That’s a great way to market that product.

Yeah. Right. But it really is around the language that they use and like kind of the very specific offer that I’m trying to have this product be sold for. Yeah.

Nice. So we have now those three areas differentiated, authentic or like with authority and specific, which very often comes down to Q&A. Now, Michel, just a call out to the audience. So like we have a lot of questions here and I will like work myself through those.

Michel asked, if generative AI can help make content clearer for machines. Do agents downgrade content that was obviously created by an LLM? We haven’t seen, I mean, it’s a great question. And it’s something that is like very constantly debated right now in SEO, AEO, GEO, whatever acronym you want to call this like industry.

We haven’t, I think there’s been like a lot of mixed evidence. Some evidence shows that, you know, AI isn’t like consistently downranking AI. But what it really cares about is like the content. It’s like quality, right?

I think overall, there’s a lot of like factors that go into quality, but overall quality, just creating a bunch of content with AI that is like kind of slop and not quality is not a great way to market towards that AI and often can be more harmful than helpful. Now, on the other side, you know, what we do work with a lot of our customers on though is there is a clear company that is downranking a lot of AI content, which is Google and SEO still matters. And so, you know, I think. Yeah.

I think that AI is a really great tool for writing and producing content, both in terms of like researching, drafting, as well as like editing. But we tell all of our customers and we highly recommend all content on the internet that is meant to market today. I should have human review on the loop, edited, righted. Maybe not 100%, but certainly not 10 or 0%.

Got it. So like from a technical point of view, I would do this. From a business point of view, I would do the following answer. Actually, there’s a business view and a technical view.

The business view is we saw XAI talk about how much money they need to train their models. We are talking in the billions of dollars. They are burning to train a model. If you spend so much money, you need to make sure that the data you feed it makes the model better.

Otherwise, it’s a complete waste. There is something in the technical space, which is called model degradation. That essentially is if you feed over and over the same information to a model, the model starts zeroing in on this. We talked on the show before about the next token prediction.

Like the next token of life is like a box of chocolate, despite the fact that this is probably the next token of life. Like it’s a nice meme, but it’s probably not true. Meaning if I spam the internet with this sentence over and over again, I will zero in on chocolate and forget everything else. Meaning the creators of those LLMs, OpenAI and Google, are extremely careful of what information they bring in.

So at the moment, you might be able to do some wins with slop. In the long run, it won’t be. It won’t be. It will hurt you.

And once you have downgraded, it’s very hard to come back. Now, there is on ask actually another question. And so like talking about the zero click experience. And does this mean companies should focus their promotion strategy on having more positive articles across the internet?

And the official website should not be actually that much focused on that. But what’s your view there? Yeah, I think a lot of things there. I would say like we, in terms of like the citation breakdown of where AI gets like essentially, you know, for each customer that we work with, the first thing that we do is we essentially run a ton of data, ton of prompts through like all the models that we care about around, you know, how is AI, where is AI getting the data to recommend products, solutions, services for the queries that we’re getting?

And so, you know, I think that’s a big part of it. Yeah. Yeah. There’s also also like these heightened heightened content is, I would say the number one thing that customers should be working on and that can have the highest driver of value for those companies.

So I think really, unless you’re in one of those very highly legislated or kind of like marketplace type industries, I think there’s never actually been a better time to build an organic growth strategy on owned content that you can control end to end on a website. And the nice thing about these content systems, right, it’s very different marketing than paid marketing. Once you stop giving a dollar to Zuckerberg, he’s probably not going to keep showing your ads after that. And it’s very much a dollar in, X out for maybe the two months that you run it.

The nice thing about these organic content programs and what you can build is it’s kind of compounding, right? And so that is generally the best strategy that we think any brand can be owning right now is building a really awesome organic marketing strategy. I wonder. I would like to bring this, like to summarize this.

This is awesome, right? This is like the compounded effect of owning your own content in a world where we go from clicks to citations. The compounded effect is huge. And that’s the direction every marketeer should think of.

Maybe I think what I would love to do, Alex, is you have built a tool, a tool which helps to. I think it’s a tool that helps to explain for a brand a little bit how they should think about citations, content, their positioning in it. And there was a lot of hype in the market around GEO tools. And I think many people get it wrong if they just think about the number of mentions.

What I would love to do, if you could give us a quick run through your tool, the one you build, and show us a little bit the setup. Yeah, absolutely. Always happy to give like a quick, quick down. We’ll try and keep this quick.

One thing I will mention here also is like, you know, I think we’re a little bit different than like a lot of the GEO companies in the space and that we from the outside and like we really brand ourselves much more like an agency, right? We don’t want to sell just like software for a company to go and do this themselves. We actually don’t think that’s a great business model. We want to like own the outcomes similar to an agency did.

Now, of course, we build all of our software in-house. And so a lot of what I’m going to show kind of here is like actually more internal technology. technology. Versus like what our partner would see because we want them taking care of their own work, right?

And not spending a bunch of time or having to spend a bunch of time on our tooling. But yeah, this is really like the overview of what we build here. When we kind of combine, you know, agents with like human experience to help these customer companies build these like organic growth engines to market and sell to AI. And the way I really think about our platform is like three main things, and I’ll try and go through them pretty quickly.

But it’s data. Analytics are, you know, on top of that data. And then based on those analytics, how do we actually improve or improvement actions and the six services that we think about? And so what we do when we onboard a company, right, is one, we first define, okay, what is like all the information about the company?

We need to be grounded in truth. What do we sell? Who are we trying to sell to? What are these ideal customers, those buyer journeys, the market that we are in and how do we like to write?

And then we gather a bunch of data around, okay, what are the types of like queries that you want to bring forth? And then how is AI, you know, responding and ranking? Yeah. Ranking different vendors, solutions, services for these queries, right?

This is for an example, staging company, Rabbit, they’re a DTC running shoe company. They also share like running apparel, definitely something they want to be mentioned for. And you can see, you know, currently getting recommended for this set of prompts, like 37% of timing rock, 8% in complexity. I won’t go too specific into all of the data, but really, you know, what we care about is who’s getting recommended, what is getting recommended, and why is it getting recommended?

We are trying at a high level. We’re trying to be able to deconstruct how and why AI is making the recommendations that it is in order to help these companies actually go from there, right? So a few of the things that we’ll look at is like sentiment, right? How is AI actually describing the brand citations?

What data is AI using to talk about this brand specifically? And then for all of the queries that they want to be the answer for, where is AI getting the data to talk about the, or essentially recommend product solutions for all of those queries? So, you know, kind of like the internet is very big, but we only care about, you know, what is the core of the internet that we actually care about, uh, and, um, optimizing around, um, and really, you know, a bunch of data and then a bunch of Linux and the way to take that into, okay, how do we actually improve? Um, and for us, it really is like six main surfaces that we think about are like the six things on the internet that matter for getting picked for AI, uh, everything from, you know, backend code, right?

This year is. The first year that there’ll be more bots on humans website than humans. Uh, we need to make sure that, you know, these bots can easily understand, uh, that code, um, as well as front of page content. Right.

Um, there’s a lot of data around how AI likes to cite content doesn’t like to cite content overall. It likes, uh, like structure. Um, I think not that to a human. Um, and so, you know, we help these companies structure their sites.

And then the third improvement action really by far the most important. And the one that we focus the most on is building that organic growth, like content engine. Right. So we have a very good idea of what do they want to rank for, uh, or what do they want to be recommended for and where are they not getting recommended?

And so we essentially take on and then building of these like content strategies. Um, of course we do a lot of like research and identification with AI, but we absolutely have like human writers in the loop to help like build out these contents. It was all staging data right now. Um, and so.

Um, but we want to build, you know, like an, an amazing growth engine that actually is, uh, one high volume, but two as important, if not more important, very high quality answers the questions talks about the companies brings in facts, context about the business versus just, you know, putting like AI slop on our partners websites. Um, and so those are really the first three improvement actions that we think about are related to our partners like website. So we can automate a lot. Um, and then we also do help on essentially like, what are the other things?

That matter. Right. Which we kind of split into like three buckets first is editorial, right? Um, if you’ve worked in SEO at all in the last like 20 years, you probably have experienced with like backlinks.

Um, and I think a lot of the strategy in the last 20 years was let’s go build a thousand backlinks and hope, you know, to really increase our visibility. Um, and a big change in like AI role today is it’s significantly more of a parallel game for every kind of like solution market service. We see there generally around 20 sources that are. Overwhelmingly trusted, maybe like 95% of the time to recommend, um, solutions for those queries.

And so, uh, a big part of like why we run all this data is we want to understand, okay, what are the editorials, um, that we need to target and help this company actually target. Like we built agents to, you know, go out and actually get them linked on these as well as, you know, any applicable social threads that are actively being trusted as well as directories. And so we take kind of this like high level approach and really for each customer that we work with, we build these like very. Uh, complex specific, you know, workflows that we were running always on with human loop to help them, uh, always be marketing with AI to other AIs.

And so this is, this is super fascinating and thanks for, for sharing. I think the, the core lesson for an executive here is you heard about geo very much about how do I, or how does my. Brand. Rank in AI.

But that’s not the question. The core question is why should an AI actually trust my content and trust my brand more than another. Is that the core summary? Yeah, I think that’s a great way to put it.

Um, I think a very simple way is, am I the answer to my customer’s questions or not? Exactly. And I am the answer to my customer’s question. Awesome.

Yes. So any content we create and you, you described this with your six. layers, which is perfect, right? You have the back end.

You saw that like me going shopping on coffee. I took screenshots. That’s super expensive. Like ordering coffee with screenshots is stupid.

Make it easy for me to be read. Front end, make it easy for me to have access, right? And make it organic content so that I have the authority, the trust, and so on and so forth. Now, let’s jump back to one of the other questions which we have.

JT asks, how does social content rank in all of this? If you have a business page with limited content, will social posts increase you relevant for LLMs? And how would an LLM interpret social posts as legit for the business or the product as an example? Very neat.

Yeah, it’s a really great question. And I think, so to back up a little bit, I think one of the more interesting questions that are going to happen in the next like five years is the question of how are these models actually get accurate data, right? It was very easy when everything on the internet was like printed by humans, whether it was accurate or not, it wasn’t specifically put there to market to like an AI. But that is like quickly shifting, right?

Right now over everything, Reddit and LinkedIn are the number one sources, of citations overall of any like webpage used by AIs to like cite solutions. And the reason is more than anything, they are trusted, they’re verified, right? To an extent. And then they’re very good like question answer type language, especially Reddit.

But I think we are going to see much more of a flood towards solutions like LinkedIn, which are a little bit harder to bot than Reddit. Reddit is a very good language for people to use. And I think it’s going to be a There’s also like一定一定一定一定一定一定一定一定一定一定一定一定一定一定一定一定一定 is on we really want to build a organic compounding things that’ll you know help like one two five years down the line um but social right now is a huge priority and so to like very specifically answer that question i would invest in social at the minimum make sure that like you have a social page on each one of the like major platforms and then it’s very explicit right um just having like a page doesn’t do a lot you should write down your products write down your services write down you know who are your pain points the jobs to be done or the problems that you solve for your customers it is kind of silly but it is very simple you need to like put very explicit language in these trusted places uh and that will greatly increase the ability of ai to one understand your brand and then to hopefully recommend your brand to potential customers talking about social um we should um actually share that we have social links obviously so if you want to follow me on linkedin you can find me on linkedin and i will be happy to help you with that and i will be happy to help you with that as well so i do post a lot of those mini videos as well as questions and answers i don’t do this to optimize for llms however you will appreciate it as much as an llm if it is specific question focused and authentic and i think that’s what brands are looking for and alex we are coming back to this point over and over again specific authentic and um question focused right now if you wonder what can be done in order to select who can select who can select who can select who can select who can select pushing out social just in order to be placed how do you as a brand think about this how do i get as a brand suddenly um a social channel if i hadn’t one before where my customers publicly talk about it about me about my brand how do i do this yeah i think it’s you know really making the shift to like building communities or building like resource hubs right um one thing that we kind of like recommend a lot is you know take i think every customer um has a lot of context about like what are the kind of questions that their potential customers like ask if you’re b2b this could be like sales recordings right if you’re b2c it could be you know the types of like uh questions you see online or in your like facebook comments whatever uh that is the best place to like start take those build content build community around them and those channels like companies that are doing well right now are making tens of millions hundred million dollars off of those signations and i search so i think it can be a gold mine if you really invest into it i can actually um bring another coffee example here the machine in the background is a decent espresso machine again i don’t have any association i just love coffee they have a very very strong community and the um ceo and um the ceo and um the ceo and um the ceo and um founder of this espresso machine um john he he speaks at some of my courses um he explains actually um how he uses that community to create more content to put it out so certainly it was novel because it was a view on how you use this machine it was authentic because it was users of that machine and um it was question based like very clearly somebody has a question this machine is the answer to to it and he managed with this formula to get up quickly in the rankings um what i love about your approach alex is you are not only focusing on the actual ranking but you focus on the workflow you are your venture funded correct so um obviously we like a venture partner wants to see a huge return here so i’m going to go ahead and i’m going to how do you how do you scale this workflow how do you scale when you say like we want to be an agency i think about agencies many many people running around costing all like a hat like a lot of headcount cost how do you in an agentic world scale that up essentially yeah i mean it’s a great question i do think you know there’s been kind of like twitter uh vibes around this recently around like what is the like actual interesting business models in the next 10 years um i think agencies were always a good business model at a small scale right the hard part was um with when you only had deterministic software they are incredibly hard to scale um but i do think we’re seeing a little bit more of like a change right now and we actually launched when we launched our product it was pure software right you pay us some money uh we run data in the software we tell you what to do we’re very quick realized there was kind of like two camps of customers one uh you know they would maybe hire somebody in-house to do the thing or take somebody like full-time make them the the owner and they would you know see like pretty awesome results or the other camp would be like oh i forgot to log in right and i think at the end of the day like nobody really wants to pay anymore for tools they don’t want to pay money to tell you what to do i think if you are building on top of like the apis the best thing that you can own is not the tooling it’s the outcome right people want to pay for outcomes they want to be paying for work done if you describe us as like an agency or like you know maybe in a genetic software it doesn’t really matter to us and i don’t really think it matters to like our customers uh our customers really at the end of the day they want to pay for outcomes they want to pay for like a plus quality work the way that we get there doesn’t quite matter um so what we’re really focused on is like we want to build the the solution the company that solves our customers problems the best we think the best way to do that is like actually on the outcomes um and so we want to like build the the product solution service that gets there uh i think a nice way to do that is like actually select the team who can select who can select who can select who can select who can select who can select I think a nice thing about AI is you can essentially productize a lot more of these flows or agency type mechanics that weren’t possible to productize in a free AI world. And I think that’s something that gets us really excited.

And this is actually what I see in the market over and over again. Big corporations think about top level. How do we go from OPEX to CAPEX? And here is Alex essentially saying in this very specific area of marketing to machines, we already offer a systemic flow of agent to agent communication, A to A, or like which we inform by a human.

So a human is steering it. That agent to agent communication, that is the new model in the new world. I’m totally in agreement here, Alex, with you. Maybe it’s a good.

Good point to tell the audience that we, Cornell is doing a bunch of workshops. The industry is changing so far. Many of you know that I have a certificate program and some of you actually are part of it. But Cornell or eCornell have thought about how can we keep up with the trend with the speed?

So we launched new workshops and one workshop is around agentic workflows. So in that workshop, it’s three hours with me. We design agents. We build agents in order to map it out for a specific use case.

Another workshop is around a GEO where we trying to do all those learnings in a specific case. So the link should be somewhere below where you can actually see you can sign up for this workshop on agentic workflows. But the key takeaway for you should be our marketing is changing. And because our marketing is changing.

We need different content and different answers in the internet. And those will need the human touch will need like like a broad distribution and so on and so forth. And that we can only do with an agentic workflow. And therefore, it’s very exciting.

Alex, what you build up. How do you? Drilling on this agentic workflow discussion. How do the skills?

Of marketing teams actually change meaning your customers are marketing teams. You coming out to them and saying, well, I I first measure you and then I have a workflow to help you. How how do your customers need to be trained to actually interact with you in the most effective way? Yeah, it’s a great question.

You know, I do think by the way, this is the question from Daniel. Thanks. Daniel. Thanks.

Great question. I think, you know, that’s why we like to or we’re starting to use more like that agency. Verbiage, honestly, because I think it is a Companies know how to think about agencies, which is okay. You know, I give context and then agency does work and reports out on outcomes.

And, you know, when we think about like our job here, we want to own that outcome. We want to build the best possible outcome. And so describing us as like an agency versus like a software tool. I think it’s the best way to do that.

If I was to sell, you know, like an agent straight up that still isn’t owning the outcome. Maybe I’m giving it a tool that might or might not actually like get to that outcome, but it’s not only the outcome. And I think for us, we want to own the outcomes. We want to be paid for the outcomes.

And we think it’s, you know, we want to make sure that our partners have full trust that we are essentially the owner of whether that is a good or good outcome versus them. Right. And I think the best way to do that is agency because when I work with these partners, they have a very clear understanding of what they’re getting. It’s let’s do onboarding.

Let’s go full context. Let’s define goals. And then ideally this. Partner of ours, you know, this customer has to do as little.

As possible. And I think that is our hope is that we do not give them a bunch of work. We can kind of just run with it. Awesome.

So we covered quite a lot of range. I trying to like I will throw you a curveball here because we covered the content specific novel. We can’t cover the workflow because there is a new approach to get the content out. What does it mean for distribution?

If I’m a if I’m a brand. I used to have my. Normal distribution brick and mortar customers and then it became Amazon. Now, if shopping suddenly happens on an agent to agent basis, how do I have to as a brand?

How do I have to rethink my distribution? And is this something that your insights can help me? Yeah, it’s a really great question. And I think we could probably talk about that for another hour.

And so I’ll give like two quick. Quick answer there. One search or like distribution is very quickly going from maybe it was a Google search bar on only that Google search bar like 15 years ago. And then, you know, social media started popping up and all that kind of stuff.

And essentially search is now everywhere. These models. Yes, you have like the actual applications, Chechuby T, Claude, whatever. But all this so these models and their question asked format are getting embedded into every place that we touch technology, right?

You’re going to have AI search in your Tesla. You’re going to have in your BMW. You’re going to have it in your phone, your laptop, every application. They use so one huge fragmentation around the search.

And then I think the second one is the like actual buyer journey is like much more from Prince where it used to be find evaluate choose and buy like four totally different mechanics and different things that you had to optimize around and like invest in. They’re all essentially kind of becoming one, right? Just like let’s put in for like the coffee example. Find me this thing like one quick prompt context on what I want agent kind of does that entire buyer journey.

You’re having this as well. So I think it’s like a very large question that lots and I don’t know. And I don’t think I can answer the whole thing. But I think those are like the two biggest trends that we were following and that should kind of be think about right now in real time.

Yes, and I am by the way, I like totally like as I said, it’s a curveball because essentially nobody really knows I published about a year ago the article that I think Amazon as a home page as an aggregator becomes. Under pressure because when I can disintermediate the top funnel of any website, then why do I need Amazon who was the aggregator now Amazon’s answer to this is because we have roofers terrible product, but like still an approach, right? And so the answer for this distribution is not yet solved. So I’m completely agree more is happening there, but I think very high level we touched on all the things which are currently happening in this.

Geo space. So first of all, it’s happening act now second of all, it’s a content question. It’s like being specific being question-based have the right technical ability. It’s a distribution question unsolved.

But yes, there will be new aggregators coming and open. AI is trying to be one. It’s a process question because if you have new content, how do I get the process out? And obviously it’s a measurement question.

As you showed in your tool. I’m sorry. Last question. I have one more.

One more. What do you think about open as open? Ai’s announcement of doing advertisements? I think it’s going to be tricky which we kind of brush on this like an hour ago, right?

But I think the reason that advertisements were so well for Google is that Google and the advertisers were aligned. They both wanted you to click on the advertisement and get. Off of Google as fast as possible. Very clear alignment systems.

That is very different from open. Ai open. Ai generally wants to keep you on. Okay.

They are not trying to send you to a bunch of links and off of Chachi BT as fast as possible. And so there’s a lot of I think questions around like ads and Chachi between that kind of thing. I think where I’m most interested is kind of this like Miss incentive align system where why am I paying Chachi BT to put an ad to send somebody to my site? Even though they are actively trying to retain that user is something that I single thing that they have to like figure out.

And I think anthropic is probably thought about this a little bit more and like long-term right now. But of the many things I think that’s like one of the more interesting questions, which we’ll see where it goes. I very much like this answer. Yes, you are absolutely right.

That’s probably as well the reason that despite big announcements. We haven’t seen a lot of. Traction in this market because it’s not that simple. With this Alex, we spend an hour.

We probably can spend two more hours or three more hours. It fascinating topic a lot of opportunities for brands to reposition themselves. And that’s the cool part. We we have a new approach towards content distribution process and a lot will happen in that space.

For all the audience. Thank you for being here. Check in with our courses check in with our websites and I see around the next few discussions are coming. Thank you.

There is also a link in the description for our exclusive exclusive exclusive exclusive exclusive exclusive exclusive exclusive exclusive exclusive exclusive exclusive exclusive exclusive exclusive